to a brick-red, or dull-brown, similar to that of the altered shale and sandstone, of which detached portions may be found adhering to the outer walls of the dyke, or even embedded in its substance. The central portion of the dyke may thus be markedly crystalline, forming what we should call a greenstone, while the outside parts, where the trap comes in contact with the adjacent rocks, are fine grained and homogeneous, so as to become a true basalt. Sometimes, too, these exterior edges are highly vesicular and amygdaloidal, detached fragments closely resembling the slag of an iron-furnace, and occasionally the dyke presents a columnar arrangement, the ends of the hexagonal or polygonal columns abutting against the sandstone or other rock on either side, and losing themselves towards the centre in the general mass of the trap. Where the strata traversed are broken and jointed, the dykes which cut them through may be seen in some places throwing out lateral veins that accommodate themselves to all the irregularities of the fissures. These minor portions exhibit for the most part the same leading features with the parent mass, and the result of the whole is a general baking of the beds, with sometimes not a little contortion, and an amount of irregularity and disturbance, apparent at once to the most inexperienced observer. (See Fig. 34.) If the reader will verify these statements by actual exploration in the field, he will probably not be long in arriving at the following conclusions: trap-dykes must once have been in a melted state, as is shown by their vesicular cavities and divergent veins; this liquid condition must have been attended with the most intense heat, as may be gathered from the burnt and baked appearance of the contiguous rocks; they have, for the most part, especially where of large size, risen from below along previously-formed dislocations--a circumstance which may be inferred from their persistency in a straight line through beds of very different resisting power, for had the liquid matter forced a way for itself, it would have squirted between the beds along the lines of least resistance, and not directly and for miles across them; and hence, trap-dykes must be regarded not as themselves the agents in dislocating and contorting a district, but merely as signs of the parent force at work below.